Closed mtytel closed 13 years ago
I realized this is ambiguous. A better tag is "run_after" or similar.
Here's an example. Job A has an hourly scheduling and Job B has the "run_after: Job A" option. When Job A completes successfully, Job B will start.
Went with a "follow_on_success" option.
Run, execute and start are ambiguous because they can either apply to the self or to the other job. Follow implies self.
Also, may consider "follow_on_failure" and "follow_regardless"* options
*probably not the best name
Our current model for job dependencies has every job on it's own schedule. When a job is scheduled to run, it checks if required jobs ran recently.
The new model for job dependencies does not have dependent jobs on a schedule. Instead have a "run_on_success" (or similarly named) option that, on completion of a required job, runs the dependent job.